Portal de Periódicos da CAPES

The Dating of Poel 11 and Hiddensee 12, Mecklenburg-West Pomerania, Germany

2015; Taylor & Francis; Volume: 45; Issue: 1 Linguagem: Inglês

10.1111/1095-9270.12139

1095-9270

Aoife Daly, Mike Belasus,

Maritime and Coastal Archaeology

For more than a decade the Hiddensee 12 and Poel 11 shipwrecks (Fig. 1), also known as the Gellen wreck and Poel cog (Förster, 2009), have been regarded as late 14th-century cargo vessels built on the south-western shores of the Baltic Sea in Mecklenburg. Hiddensee 12 was discovered in 1996 during a survey by the regional society for underwater archaeology on the south-west coast of the island of Hiddensee. Soon after it was located, a few medieval artefacts found nearby, the presence of axe marks on all visible timbers, and the use of double clenched nails to connect the planks in a clinker construction style, quickly led to the assumption that the ship was of late medieval, probably 13th–14th century, date (Förster, 2009: 79–80). This was later confirmed by dendrochronology, which resulted in a dating of after AD 1378. According to this analysis the trees were cut most likely in Mecklenburg. Other possibilities were North Poland or Brandenburg (Heußner, 2004a). Poel 11 was discovered in 1999 on the west coast of the island of Poel. Because of the resemblance of its technical features to those of Hiddensee 12, a similar date was postulated by the excavator and later confirmed using dendrochronology (Förster, 2000: 225–6). For this analysis, 17 of the ship's treenails, each containing up to 80 rings, were sampled. No other timbers were sampled. The results gave a felling date for the trees after AD 1369. An origin for the timber in Mecklenburg or southern Finland seemed to be most likely (Heußner, 2004b). Both ships were built from conifer wood, which is uncommon in medieval ships of the south-west Baltic Sea area where oak was the predominant building timber during this period. Furthermore, both ships are extraordinarily large for the medieval south-west Baltic Sea and bear a number of technical features that are both unusual in medieval shipbuilding and in the combinations noted (Belasus, 2015). Nevertheless, both ships were regarded as late medieval cargo vessels of the new 'Baltic cog' type, even though they were built using the clinker technique (Lüth and Förster, 1999). Therefore this classification had nothing in common with the generally accepted archaeological-technical definition of term cog (Crumlin-Pedersen, 2000; Dammann, 2000). Instead, it refers to the work of historian Paul Heinsius, who defined regional sub-categories of cogs, on the basis of town seals, in his book Das Schiff der hansischen Frühzeit (1956). As part of a project to research the medieval ship finds of the south-west Baltic Sea, carried out from 2008 until 2011 by the German Archaeological Institute's Roman Germanic Commission in cooperation with the Heritage Agency of Mecklenburg-West Pomerania and the German Maritime Museum, a new dendrochronological analysis of both ships was carried out (Belasus, 2012). The aim was to get precise information on the origin of the timbers used for these unusual medieval clinker vessels, and to explain the features that deviate from those commonly recorded on medieval ship finds of the south-west Baltic Sea (Belasus, 2014; 2015). The 25–27 May 2010, samples were taken from timbers from the Poel 11 and Hiddensee 12 wrecks, for dendrochronological analysis. Equipment was transported to the storage sites of the wrecks so that the analysis could be carried out at the location, allowing the samples to be returned immediately to their water-filled tanks. Later, in September 2010, additional samples from Hiddensee 12 were taken, and analysed at dendro.dk in Denmark. In October 2010, additional measurements from the Poel 11 wreck were kindly sent by Karl-Uwe Heußner, of the Deutsches Archäologisches Institut. The results of the dendrochronological analysis of these two wrecks were described in a report in January 2011 (Daly, 2011), including new dates for the Poel 11. In 2015, Daly was able to look again at the tree-ring data for the Hiddensee 12 wreck, and as additional chronologies have recently become available, this ship is now also dendrochronologically dated. This note is intended to bring these results to a worldwide audience. A description of dendrochronological methods can be found in a wide variety of publications and is not reiterated here (see for example Baillie, 1982; Hillam, 1998, Daly, 2007). The samples were taken using a handsaw and the cross section of the timbers were carefully pared with a razor blade and then rubbed with chalk to allow the tree-rings to be clearly visible. Measurements were made using a stereo microscope and measuring stage designed to run with DENDRO (Tyers, 1997). The DENDRO program is also used for cross-matching, averaging and managing the tree-ring data. For the calculation of the t-value ('t-test') CROS (Baillie and Pilcher, 1973) is used. A total of 16 timbers and 15 treenails from the Poel 11 ship have been examined. Six samples were taken from ship frames by coring, but only one of these is dated. All ten samples sawn from planks are dated. Seven treenails, three of which were analysed by Heußner, are also dated. Two of the frames are of Picea sp. (spruce), while the remainder of the material analysed is Pinus sp. (pine). Table 1 shows the degree of internal correlation (t-value) between the tree-ring curves from the 18 dated samples. The tree-ring curves from two planks (planks 130 (Z0330079) and 140 (Z033102A)) are very similar to each other, leading to the suggestion that the planks might be from the same tree. These two were averaged to create a single tree-ring curve (Z0331509). The tree-ring curves were averaged to form a site chronology of all dated tree-ring curves (Z033M001) (Table 2). The chronological position of the 18 dated tree-ring curves is shown in Figure 2. The presence of bark edge could not be discerned with certainty on the samples. However, on many of the samples, the outermost preserved ring was formed in AD 1770 to 1773. On this basis, we can conclude that the trees used to construct the ship were felled in AD 1773 or shortly after. As can be seen from the table of t-values generated by correlating the mean curve for the ship and master chronologies for pine from around Scandinavia (Table 3) the trees used to build Poel 11 probably grew in the region of south-west Finland. Twelve timbers from the Hiddensee 12 ship have been examined. Generally, the timbers contain relatively few rings, making cross-matching difficult. It is, however, possible to cross-match tree-ring curves from eight of the samples through internal comparison and by comparison with large regional pine chronologies (Table 3). The dating results obtained for these samples are also shown in Figure 2. None of the timbers have bark edge preserved. The most recent, outermost tree-ring preserved is on sample Z0322032, which is from a tree that was felled after AD 1804. The dating of outermost rings on many of the other samples from the Hiddensee 12 fall shortly before this date, suggesting that not many of the outer rings of the trees were trimmed in fashioning timbers for the ship. The ship was built probably in 1804 or shortly after. An average of these eight series has been made (Z032M001) that is 113 years long. The mean tree-ring indices for the Hiddensee 12 ship are shown in Table 4. The correlation between the average for the Hiddensee 12 ship (Z032M001) and diverse chronologies for pine is shown in Table 3. As with the Poel 11, the Hiddensee 12 tree-ring series is most closely cross-matched with chronologies from Finland. The two ship chronologies achieve a t-value of 4.36 when compared at their dated position. This might indicate that, while the two are made from pines that probably grew in the wider Finnish region, they might come from different areas in that region. These new results are crucial for Baltic ship archaeology. The interpretation of both ships as large medieval cargo vessels from the south-west Baltic caused them to be attributed to the sub-category of 'Baltic cogs'. This has served to isolate them from their archaeological and historical context and precluded the satisfactory explanation of their specific features. Their appearance in some research literature has contributed to a distortion of the image of medieval shipbuilding (for example Meier, 2006: 38; Hammel-Kiesow, 2009: 126; Krause, 2010: 95–7). In fact, both ships have caused considerable confusion among scholars of medieval ship archaeology. The new dendrochronology results make it possible to place the ships in an early modern context that allows their particular construction features to be explained as products of restricted development in peasant shipbuilding and seafaring during an economic boom period in Swedish Finland (Belasus, 2015). The analysis has been aided by input from Cathy Tyers, Sheffield University; Anne Crone, AOC Archaeology Group, Edinburgh; Maris Zunde, University of Latvia; Karl-Uwe Heußner, Deutsches Archäologisches Institut; Pentti Zetterberg, University of Joensuu; Kristof Haneca, Flanders Heritage Agency; and Alar Läänelaid, University of Tartu.